Captcha code recognition method and system based on improved YOLO9000 algorithm

By improving the combination of the YOLO9000 algorithm and the DBN model, the problems of low accuracy and slow speed in CAPTCHA recognition were solved, achieving fast and accurate Chinese character recognition and enhancing the model's generalization ability and recognition efficiency.

CN114529911BActive Publication Date: 2025-11-04INSPUR TIANYUAN COMM INFORMATION SYST CO LTD
2 Cites 0 Cited by

Patent Information

Application Number
CN202111634916.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-29
Publication Date
2025-11-04
Estimated Expiration
2041-12-29

AI Technical Summary

Technical Problem

Existing technologies for CAPTCHA recognition suffer from low accuracy, slow speed, and model overfitting. In particular, when dealing with CAPTCHAs containing Chinese characters, traditional methods struggle to extract effective features, while deep learning models suffer from complex parameter tuning and large storage requirements.

Method used

An improved YOLO9000 algorithm is adopted. By reducing the convolutional layers of the bottleneck structure, and combining the cross-entropy loss function and the average difference similarity function, a YOLO model is constructed. A deep belief network (DBN) is then used for feature extraction and recognition. The recognition results of the two models are fused using a linear confidence accumulation method.

Benefits of technology

It improves the accuracy and speed of CAPTCHA recognition, reduces the model's storage space and computational complexity, enhances the model's generalization performance, and avoids overfitting.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The application discloses a Captcha recognition method and system based on an improved YOLO9000 algorithm, belongs to the technical field of Captcha recognition, and aims to solve the technical problem of how to quickly and accurately realize Captcha recognition. The method comprises the following steps: collecting Captcha pictures, wherein Captchas in the Captcha pictures are composed of Chinese characters, English and numbers; labeling Chinese characters in the Captchas and performing image transformation processing on the Captcha pictures; for a YOLO9000 model, the convolution of a bottleneck structure is reduced, the convolution layer is improved, a cross-entropy loss function and an average difference similarity function are used as a loss function, and a Captcha recognition model is constructed based on the improved YOLO9000 model; a Captcha recognition model is constructed based on a DBN model; the YOLO model and the DBN model are trained based on the preprocessed training samples; and the classification results are fused by adopting a linear credibility accumulation mode.
Need to check novelty before this filing date? Find Prior Art